Google Analytics interpretation, a true story!

Consider these statistics

Initial conclusions?

Wow, what a low bounce rate! And what a low % of search engine traffic! Lets seo (optimize the site to get more (organic) traffic from search engines):

1) replace the table based layout with css based template (through lack of a template, every little site-wide amendment had to be made seperately for every single page and the layout lacked consistancy)
2) make sure all pages without text are removed (most primary navigation buttons directed visitors to a page with secondary navigation but without any content. All those empty pages were also indexed by search engines)
3) redirect all old .html urls with a 301 to new corresponding .php urls.
4) optimize html emphasis (h1 tags were temporaly omitted for the sake of a quick re-launch); a keyword selection tool was used!
5) add personalized error pages (previously, error 404 not found used to appear without site layout thus no possibility of navigating through the domain)
6) 301 redirect non-www version to www domain

Consider these statistics from a week later after a complete re-design of the site

Initial reaction

WHAT HAPPENED WITH THE BOUNCE RATE AND DIRECT TRAFFIC!
I mean, direct traffic is supposed to come from:
1) people that directly enter the url in the browser address field,
2) people that have navigated to the site using their bookmarks.

I cannot believe that re-building the site may have caused a sudden drop for either of these two criteria of direct traffic NOR do I understand why the bounce rate has significantly increased, while the content has remained on the site practically unchanged.

It is implied that there may be other criteria for visits to be specified as "direct traffic" and for visits to be registered as bounces.
WAS THE SEO A MISTAKE?

Reconsideration

old scenario

Many visitors, including direct traffic, used to land on a page without content. In order to see their desired content, they gave the site one more chance and clicked through to a page either with or without content and may have left. They were counted as a visit, thus no bounce.

current scenario

In the new situation, visitors arrive directly on a page with content. If they leave now, they won't be counted as a visitor. Thus the increase of the bounce rate. And since direct traffic used to be the main source, it is most strongly effected by this seemingly declining statistic.

Final conclusions

SEO WAS SUCCESSFUL; organic search is now the leading traffic source!

The original low bounce rate was "artificially" created through the fact that most people landed on a page with only navigation, no content. They clicked once to find some content and were registered as a visitor, regardless any following actions like leaving the site in particular.

The "new" bounce rate is a much more realistic indication for the site's stickyness than the "old" one. It's also still very respectable!

The "no content" situation evidently caused misleading statistics.

Still, nice benchmarks.